The aim of this research is to design an autonomous real-time air quality monitoring gas sensor for smart green building. The smart green building contains two features, one is energy efficiency of heating, ventilation and air conditioning control system (HVAC), and the other one is a healthy and comfortable environment for the users. These can be realized by real-time indoor air quality monitoring using the gas sensor. Air quality monitoring nowadays is mainly limited to detect smoke and prevent fire. However, harmful substances can cause a risk to public health. The gas sensor can be deployed to detect such substances, monitor the air quality, and enable smart HVAC control. The large deployment of these gas sensors in buildings is currently difficult to achieve because of: 1) the larg...[ Read more ]

The aim of this research is to design an autonomous real-time air quality monitoring gas sensor for smart green building. The smart green building contains two features, one is energy efficiency of heating, ventilation and air conditioning control system (HVAC), and the other one is a healthy and comfortable environment for the users. These can be realized by real-time indoor air quality monitoring using the gas sensor. Air quality monitoring nowadays is mainly limited to detect smoke and prevent fire. However, harmful substances can cause a risk to public health. The gas sensor can be deployed to detect such substances, monitor the air quality, and enable smart HVAC control. The large deployment of these gas sensors in buildings is currently difficult to achieve because of: 1) the large power dissipation or consumption of these sensors, which would require frequent battery replacement; 2) Low CMOS compatibility, which may require special integration techniques to achieve an integrated wireless sensor node; 3) slow response time, which may need a long time to detect risk gases; 4) low selectivity, which may be difficult to separate one particular gas with other gas classes. In addition, the existing air quality monitoring mainly relies on the connection of the mobile phone, which still needs human interaction, and not suitable for autonomous operation. The objective of this thesis is to investigate, design and fabricate smart ultra-low power autonomous gas sensor for real-time air quality monitoring, and ultra-low power consumption will enable large-scale deployment with low maintenance cost. In contrast to existing gas sensors, the proposed sensor will not rely on the software of mobile phones. Instead, it will enable a gas event detection. With the early detection, more information in the transient response region will be extracted and quantized for air contaminants recognition once a gas event is detected, instead of waiting for a long period due to slow response time. The reduced quantization tasks of only feature information will further reduce the power and the data bandwidth for wireless transmission. In addition, new gas-sensitive FETs (Gas-FETs) is fabricated in the standard CMOS process, which can operate at the sub-threshold region, and readout through dynamic current integration. Thanks to the work function gas sensing mechanism, the response time and sensing area can be reduced compared with previously reported resistive gas sensors in the standard CMOS process. Thus a large Gas-FET array can be integrated with processing circuits on the same chip, which enables large array sensor response pattern collection for gas classification purpose. To improve the selectivity of the array when there is interfering gas, a selective electronic nose is designed and fabricated in the standard CMOS process, with multiple sensing mechanisms and operation conditions modulation. The feature vectors extracted from the electronic nose are demonstrated in gas classification. Meanwhile, flow sensing function is integrated on the same chip. The known gas class from the recognition of the electronic nose will help the correct flow calibration in the particular gas environment, eliminating environment gas change induced error in the flow monitoring. The large array will also help the flow pattern recognition with different flow input angles. Six different angles with minimum 30°have been identified in the measurement.